Figure 4 from Late-Stage Outcomes as Surrogates for Mortality in Cancer Screening Trials: A Systematic Review and Meta-analysis

Exploratory analysis of the effect of the timing of the reporting of late-stage incidence on the association with the mortality outcome. The figure includes trials that reported their late-stage outcomes at least twice [at intervention phase (triangles), earlier than mortality (circles), and/or at mortality (squares)]. Each trial is then represented by two or up to three points connected by a straight line drawn at the main estimate of the screening effect on cancer-specific mortality; this line shows how the estimates of the screening effect on the late-stage incidence within a specific trial changed over time. Size of the points corresponds to precision of the mortality result (inverse variance of the estimate). Colors denote types of cancer (please see the legend).